Improving the seismic isolation for the AEI 10 m prototype

Abstract

The first detection of gravitational waves (GWs) on September 14, 2015 as well as the subsequent observations brought forth a new era in astronomy, which offers new insights into the nature of the universe. Future GW observatory networks, with even higher sensitivity, will make GW observations a routine occurrence. Test facilities, such as the AEI 10m prototype, provide an environment for developing novel technologies for these future detectors. The AEI 10m prototype offers a low noise environment, ideal for performing high precision physics, such as measurements at and below the standard quantum limit (SQL) of interferometry, a regime so far largely unexplored. At low frequencies, seismic noise is a major limitation for ground-based GW detectors, where isolation from environmental vibration is crucial. This thesis presents the performance analysis of and improvements to the AEI 10m prototype Seismic Attenuation System (AEI-SAS). The fundamental resonance frequencies of optic suspensions of the AEI 10m prototype are between 0.5Hz and 30Hz, a frequency band in which efficient vibration reduction is essential. Three individual AEI-SAS units provide vibration isolation to the three optical tables which support the suspensions and other optical mechanics of the AEI 10m prototype. This system decouples the payload from ground vibration in six degrees of freedom by means of low fundamental resonance frequencies and a low active-bandwidth isolation scheme. The scheme employs Inverted-Pendulum (IP)-legs to provide horizontal vibration attenuation and Geometric Anti-Spring (GAS)-filters to provide vertical isolation. Two AEI-SAS units were assembled and implemented in the AEI 10m prototype. A detailed analysis of these units revealed their mechanical limitations. Internal resonances, in particular, severely reduced their isolation performance above 8Hz. As a solution, several design improvements, such as an additional well damped isolation stage, inertial damping structures, a new IP-leg design, and a stiffer GAS-filter support structure were suggested in the scope of this work. The former two improvements were implemented in the first two AEI-SAS units. The third AEI-SAS unit employs all the improvements listed above. A revised GAS-filter tuning technique further improved its high vertical isolation performance at frequencies above 30Hz. Performance tests of the first two AEI-SAS units now show improvement in passive isolation from ground motion by a factor of about 400 in the horizontal direction at 4Hz and in the vertical direction at 9Hz. Internal resonances limit the isolation performance above 30Hz. The amplitude of the internal resonances, however, is reduced by up to a factor of 75 by means of the inertial damping structures. The third AEI-SAS unit's internal modes have been stiffened to a lowest resonance frequency of 54Hz. The AEI-SAS, therefore, now provides substantial passive isolation at all the fundamental mirror suspension resonances. The improvement in passive performance allow the implementation of an enhanced active low frequency isolation, since internal resonances strongly influence the control loop design. Active isolation reduces the payload motion efficiently in the frequency band around the AEI-SAS's fundamental resonances. Future upgrades, such as the implementation of horizontal low-noise table top sensors, will further enhance the active performance of all three units. The improvements to the AEI-SAS, developed in the scope of this thesis, have potential benefits for GW detectors and other experiments requiring low seismic noise conditions. The studies presented here influence and guide the design of future vibration isolation systems for high precision measurements in general and ground-based gravitational wave observatories in particular.